
Design a compute capacity planning dashboard
You'll have a wireframe or mockup of a dashboard showing projected GPU utilization, cost per job, and failure risk for the next quarter.
What you'll make
AI Capacity Planning Dashboard
- Header: demand projection over 12 months
- Central panel: utilization vs capacity for each constraint type
- Side panel: project priority ranking based on resource efficiency
- Bottom alerts: constraint violations and escalation thresholds
Upload a wireframe or mockup showing the dashboard layout with actual constraint data and project priorities from your analysis
The 3 parts
Each part ends with a check you can see.
Find your binding constraints
Test which infrastructure layer will actually limit your AI scaling first
Done when
You have identified the specific constraint that will limit AI workload scaling in your environment with quantified capacity numbers
Rank projects by resource efficiency
Decide which AI projects get priority when infrastructure capacity forces trade-offs
Done when
You have a ranked project list with resource requirements and business justification for the priority order when capacity is limited
Design the capacity dashboard
Create a wireframe that surfaces infrastructure constraints and project priorities before they block your AI roadmap
Done when
You have a dashboard wireframe that shows demand vs constraints with project ranking and alert thresholds based on your actual constraint analysis
Capstone: you put together your AI Capacity Planning Dashboard.
Where this helps
An elective: a single lesson that sits outside rtwo's job certificates and doesn't count toward one.
About this lesson
What you'll be able to do
- Design a dashboard layout that shows your AI compute demand against real infrastructure constraints
- Make capacity allocation decisions based on business priorities when infrastructure limits force trade-offs
- Build constraint models that account for power, cooling, and network limits beyond GPU availability
- Have a tested approach for identifying which infrastructure bottlenecks will hit first in your environment
Who it's for
- Infrastructure teams planning AI deployments who need to justify resource allocation
- Engineering managers who have to decide which AI projects get priority when capacity is limited
- Anyone who needs to model real infrastructure constraints rather than optimistic demand projections
What it won't cover
- Detailed GPU architecture or chip-level optimization
- Data center real estate investment strategies
- Production MLOps deployment patterns
- Specific vendor pricing or contract terms
Skills you'll practise
- Infrastructure constraint identification
- Resource allocation decision-making
- Constraint visualization design
- Capacity modeling under real limits